Glioma represents the single most frequent primary brain malignancy. The most advanced form of glioma, also the most invasive, is glioblastoma or GBM that comprises 50-60% of all gliomas. The median survival for GBM patients is less than a year. Only about 20% of patients respond to therapy and live for more than 1.5 years following initial diagnosis. Intensive research during the last two decades has demonstrated that cancer progresses with an accumulation of genetic and molecular alterations that enables cells to acquire at least six capabilities (Hanahan and Weinberg, Cell 100:57-70, 2000). Evidence from in vitro and in vivo model systems has been accumulated that identifies some of those events at gene expression level. Animal models including transgenic and knockout models have been used to test the importance of some of the identified molecular events in in vivo setting (van Dyke and Jacks, 2002). For example, expression of EGFR (Holland et al., 1998), PDGF (Dai et al., 2002), K-ras in combination with akt (Holland et al., 2000), have been shown to induce gliomas in transgenic models, which has been highly valuable to sort out what the key events are in tumor formation and tumor phenotypes. Some models, such as the glial-specific mouse model (the RCAS/tv-a system) developed by Holland and Varmus (Holland et al., 1998) permit transgenic expression of multiple genes in one experimental setting thus the effect of gene combination can be evaluated. Using genomics and informatics approaches, we have profiled gene expression in different grades of gliomas and identified genes, such as insulin-like growth factor binding protein 2 (IGFBP2), that are uniquely overexpressed in GBM (Fuller et al., 1999). Our in intro experiments showed that IGFBP2 overexpressing cells was more invasive. IGFBP2 was also associated with shortened survival in GBMs (Sallinen et al., 2000, and our unpublished data). These studies led us to hypothesize that IGFBP2 is a key molecule that contributes to invasion of GBM cells and short survival of GMB patients. We also hypothesize that IGFBP2 overexpression plays a positive role in glioma progression to GBMs. In this project, we propose to test the hypotheses using the RCAS/tv-a mouse model. We further hypothesize that GBMs developed in mice are homologous to human GBMs and this can be evaluated by comparative gene expression profiling as well as focused set of genes. The three specific aims are: (1) To test whether IGFBP2 enhances GBM development shortens survival of GBM mice, and increases GBM cell invasion in vivo. (2) To examine whether IGFBP2 plays an important role in progression from low-grade glioma to GBM. (3) To characterize mouse gliomas using gene expression profiling. Through this program, we want to enrich our knowledge of IGFBP2 pathways that are likely important for glioma development and progression; we also want to use genomics approach to evaluate whether mouse gliomas faithfully represent human gliomas at gene expression levels. [unreadable] [unreadable]